Effects of Vegetable–Fish Co-Culture on CH₄ and N₂O Emissions from an Aquaculture Pond

Abstract

Freshwater aquaculture is an important source of greenhouse gas (GHG) emissions. GHG emissions are expected to lead to global warming and climate change. A reduction in GHG emissions is urgently required for the sustainable development of freshwater aquaculture. In this study, a laboratory-scale experiment was conducted to analyze the effects of a vegetable–fish co-culture on CH₄ and N₂O emissions from a freshwater aquaculture pond. The results show that the co-culturing of yellow catfish with pak choi (PC-F) or water spinach (WS-F) significantly reduced the N₂O emission from the aquaculture pond by 60.20% and 67.71%, respectively, as compared with a yellow catfish monoculture (F). However, the co-culture of these two vegetables did not affect the level of CH₄ emissions. The reduction in N₂O emissions was primarily attributed to the decrease in the concentration of N₂O and NO₃⁻ in the water. The overall global warming potential (GWP) of CH₄ and N₂O was significantly reduced by 19.1% with PC-F compared to F, but it did not significantly differ between WS-F and F. PC and WS cultivation improved the food yield by 1555.52% and 419.95% compared to F, respectively. Consequently, the GHG emissions intensity (GHGI) under PC-F and WS-F decreased by 96.15% and 80.77% compared to F, respectively. Altogether, the results highlight that a vegetable–fish co-culture is likely an efficient system for mitigating GWP per unit of food yield in freshwater aquaculture ponds. These results can provide a reference for the mitigation of GHG emissions from freshwater aquaculture.

1. Introduction

Aquaculture, which accounts for nearly half of all fish production, plays an important role in the protein food supply. Aquaculture ponds are widely used fish-culture systems used around the world. In order to achieve high fish yields and greater economic benefits, intensive aquaculture with a large amount of artificial pelleted feed is widely practiced in aquaculture ponds [1]. However, intensive aquaculture-induced abundant residual carbon and nitrogen are accumulated in the water and sediment of ponds, which leads to CH₄ and N₂O emissions [2,3]. A previous study estimated that the total CH₄ emissions produced from global freshwater aquaculture ponds was 14Tg CH₄ yr⁻¹, of which almost half was accounted for by rice cultivation [3]. The amount of N₂O emissions from aquaculture was calculated to be 93 Gg in 2009 and is predicted to increase to 383 Gg by 2030, which would account for 5.72% of anthropogenic N₂O emissions [4]. Therefore, reducing the CH₄ and N₂O emissions from aquaculture ponds is an urgent requirement for the sustainable development of the aquaculture industry.

Co-culturing fish with vegetable crops is a recommended strategy to improve nutrient-use efficiency and to reduce the negative environmental effects induced by residual nutrients [5,6]. Vegetables can be co-cultured with fish using aquaponic systems to treat the aquaculture wastewater or directly in the pond with a floating bed [6,7]. Many previous studies have tested the impact of these co-culture systems on nutrient element (e.g., N and P) recycling processes and nutrient-use efficiency [8,9]. In recent decades, some studies have investigated the impact of aquaponic systems on N₂O emissions from fish aquaculture. In the studies, co-culture with aquaponic systems significantly reduced the N₂O emissions. Previous studies have shown that N₂O emissions from aquaponics accounted for 0.72–2.53% of their total nitrogen input [10,11,12,13], which is similar to the N₂O conversion ratio of 0.18–2.46% in aquaculture ponds [14,15,16]. Hu et al. [12] reported that aquaponics could also be an important source of anthropogenic N₂O emissions. However, previous studies on vegetable–fish co-culture have primarily focused on aquaponic systems. The effect of integrating fish cultures with floating bed vegetable systems on CH₄ and N₂O emissions from aquaculture ponds has rarely been studied.

China is one of the primary aquaculture nations, and it is responsible for 58% of global aquaculture production [17]. Ponds are the dominant culture system for freshwater aquatic animals. The area and production of aquaculture ponds reached 264.5 × 10⁴ ha and 2230 × 10⁴ tons in 2019, accounting for 51.7% and 74.0% of the total area and production of freshwater aquaculture in 2019, respectively [18]. Co-culturing with vegetables in aquaculture ponds is an ecological measure that can be used to reduce water eutrophication and increase a farmer’s income [19]. We hypothesize that vegetable cultivation with the floating bed system can help with the uptake the of nitrogen in water and mitigate CH₄ and N₂O emissions. In this study, a laboratory-scale experiment was conducted to determine: (1) the effect of vegetable–fish co-culture on CH₄ and N₂O emissions (i.e., CH₄ and N₂O dynamics, concentrations of CH₄ and N₂O in the water and bottom soil, GWP, and yield-scaled CH₄ and N₂O emissions), chemical properties in water and soil (total nitrogen (TN), available nitrogen and total organic carbon (TOC)), and the yield of vegetables and fish; and (2) how vegetables mitigate CH₄ and N₂O emissions by influencing water chemical properties.

2. Materials and Methods

2.1. Site Description

A laboratory-scale co-culture system was implemented in the experimental farm of China National Rice Research Institute (longitude: 119°57′, altitude: 30°03′, and elevation: 21 m) located in Hangzhou, Zhejiang Province, in 2019. The experiment site is located on a plain with a subtropical monsoon climate (annual average temperature of 17 °C and annual average precipitation of 71.6 mm). The soil was collected from an adjacent aquaculture pond by continuously culturing fish over 10 years, with a value of alkalinity (pH) of 5.82. The soil contained 1.45 g kg⁻¹ TN, 59.18 mg kg⁻¹ ammonia nitrogen (NH₄⁺), 56.16 mg kg⁻¹ nitrate (NO₃⁻), 0.60 g kg⁻¹ total phosphorus (TP), and 22.50 g kg⁻¹ organic matter.

2.2. Experimental Set Up

The laboratory-scale co-culture system was used to test the impact of vegetable–fish co-cultures on CH₄ and N₂O emissions from aquaculture ponds. Plastic tanks (length 1.75 m, width 1.25 m, and height 1 m) were filled with 20 cm of bottom soil and 65 cm of water. This experiment included three treatments: pak choi (Brassica rapa var. chinensis (L.) Hanelt)–yellow catfish co-culture, water spinach (Ipomoea aquatica Forsk)–yellow catfish co-culture, and yellow catfish monoculture as the control. Each treatment contained three replicates. They were arranged in nine plastic tanks.

The vegetables were planted on the floating beds in the tanks (Figure 1). The floating bed was made using an extruded polystyrene board perforated plate, which was 0.4 m in length, 0.4 m in width, and 0.04 m in height. Nine holes (diameter 0.03 m) were evenly distributed in the floating bed with an interval of 0.08 m. The pak choi and water spinach were transplanted in the cultivation basket after being grown on a seedbed until they reached 10 cm in height on 21 June 2019. One plant was placed into each of the cultivation baskets with a sponge. They were then planted into the hole in the floating bed. Four plates of floating beds were closely arranged in one tank. No chemical fertilizer, pesticide, or herbicide was applied for vegetable cultivation during the experimental period.

The yellow catfish fingerlings were stocked in the tank at a density of 80 fingerlings per tank (4–5 g per fingerling). After fish stocking, we aerated the water in the tank during the hours of 9:00–11:00, 14:00–18:00, and 23:00–6:00, excluding the sampling time. Commercial pellet feeds were used to feed the fish twice daily and the amount of feed was adjusted according to the change in the weight of the fish and the weather. The vegetables were harvested at an interval of four weeks and the last harvest was conducted on 23 October 2019. The yellow catfish were harvested on 24 October 2019.

2.3. Sampling and Measurement Methods

We used the static closed chamber method with steel support [15,20] to collect CH₄ and N₂O samples once per week. The sample collection chamber was made from stainless steel materials, and the size of the chamber was 50 cm (length) × 50 cm (width) × 50 cm (height). A fan was placed on the inner top of the static chamber to ensure complete gas mixing in the process of sampling. Two holes were cut at the top and side of the static chamber to determine the gas temperature and to collect gas samples in the chamber with a thermometer sensor and PU pipe. The steel frame was fixed as the base of the static chamber at the sampling site after rice transformation. Four gas samples from each pot were obtained by using an automatic GHG sampler at 10 min intervals (0, 10, 20, and 30 min after chamber closure) during the hours of 8:00–10:00, and gas samples were collected once a week from the third day after the floating bed was placed to the day before the last vegetables were harvested. During the study period, gas samples were collected 18 times.

The concentrations of CH₄ and N₂O in the gas samples were analyzed by performing gas chromatography (GC 2010, Shimadzu, Kyoto, Japan). CH₄ and N₂O levels were determined by using a flame ionization detector (FID) and ⁶³Ni electron capture detector (ECD), operating at 300 °C and 350 °C, respectively. N₂ and a mixture of Ar/CH₄ (95%/5%) were used as the carrier gases, respectively. The detectors were calibrated daily using standard gas mixtures produced by Dalian Special Gas. The certified concentrations of CH₄ and N₂O were 9.6 ppm and 3 ppm, respectively. The CH₄ and N₂O fluxes (F_CH4 or F_N2O) were calculated as follows:

F = ρ × V/A × P/P₀ × dC/dt × 273/(273 + T) × 60

(1)

where F is the CH₄ (mg m⁻² h⁻¹) or N₂O fluxes (μg m⁻² h⁻¹), ρ is the density of CH₄ or N₂O at the standard state (μg m⁻³), V is the volume of the chamber (m³), A is the basal area of the static chamber (m²), P represents the atmospheric pressure in the static chamber (MPa), P₀ represents the standard atmosphere pressure (MPa), dC/dt is the CH₄ or N₂O change in the chamber (10⁻⁶ min⁻¹), and T is the mean temperature in the chamber.

To measure concentrations of CH₄ and N₂O in the water and bottom soil, 10 mL of water and soil solution samples were obtained with a 60 mL vacuum syringe via a PU pipe at a 10 cm depth and with a MaxiRhizon-10 sampler (length, 90 mm; diameter, 4.5 mm). The samples (10 mL) were immediately injected into 20 mL headspace bottles with a syringe. All of the samples were equilibrated by filling them with high-purity N₂ for 5 min. After heavy shaking by hand, 5 mL of gas from the headspace bottles was collected and analyzed by gas chromatography for CH₄ and N₂O [21]. The water and soil solution samples were collected once a month from the third day after the floating bed was placed. The CH₄ and N₂O concentration (C_CH4/N2O) in the water and soil solution was calculated using:

C_CH4/N2O = (m × G_V)/(G₁ × M_V)

(2)

where m is the mixing ratio as the phase (ppmv), G_V is the volume of the gas headspace of the bottle, G₁ is the volume of liquid in the bottle, and M_V is the gas volume of an ideal gas (CH₄, 24.78 L mol⁻¹, at 25 °C; NO₂, 37.1 L mol⁻¹, at 25 °C).

Water samples were collected simultaneously by carrying out gas sampling. Soil samples were collected simultaneously with vegetable harvest at a depth of 0–10 cm. The samples of fish and vegetables were collected at the time of harvest and the yields of fish and vegetables were recorded.

The contents of TN, NH₄⁺, NO₃⁻, nitrite N (NO₂⁻), TOC, biological oxygen demand (BOD), and chemical oxygen demand (COD) in the water were analyzed using standard methods (SEPA, 2022). The concentrations of TN, NH₄⁺, NO₃⁻, soil organic carbon (SOC), dissolved organic carbon (DOC), microbial biomass carbon (MBC), dissolved organic nitrogen (DON), and microbial biomass nitrogen (MBN) were analyzed using standard methods [22]. The concentrations of nitrogen in rice and fish were determined by using the Kjeldahl method. The concentrations of dissolved oxygen (DO) and pH were measured by using a portable apparatus (Mettler Toledo, Seven2Go Pro S9; Mettler Toledo SG2) in situ.

The GWPs (kg CO₂-eq ha⁻¹) of CH₄ and N₂O per unit mass were effectively equal to 25 and 298 times that of CO₂, respectively. GHGI (kg CO₂-eq kg⁻¹) represents the seasonal GWP per unit yield of fish and vegetables. The formulas are as follows:

GWP = CH₄ × 25 + N₂O × 298

(3)

GHGI = (CH₄ × 25 + N₂O × 298)/(fish yield + vegetables yield)

(4)

2.4. Data Analysis

A one-way ANOVA was used to analyze the difference between the means of the three treatments in the experiment. All of the analyses were performed with SPSS 18.0. Differences between treatments were considered significant at p < 0.05.

3. Results

3.1. CH₄ and N₂O Emissions

The vegetable–fish co-culture did not significantly reduce the CH₄ flux rate as compared with F (Figure 2a). The average CH₄ flux rate decreased by 8.14% and increased by 20.43% when using PC-F and WS-F compared to using F alone; however, the discrepancy was not significant (

Table A1. The mean flux rates of CH₄ and N₂O in all the treatments.

	CH4 (mg m−2 h−1)	N2O (μg m−2 h−1)
PC-F	1.08 ± 0.26 a	10.48 ± 2.07 b
WS-F	1.42 ± 0.36 a	8.50 ± 0.91 c
F	1.18 ± 0.21 a	26.33 ± 4.30 a

Different letter indicate statistical significance at the 0.05 level.

).

The flux rate of N₂O for F showed the following two flux peaks: the first peak was 141.33 μg m⁻² h⁻¹ on 8th July and the second peak was 80.62 μg m⁻² h⁻¹ on 14th October (Figure 2b). Conversely, for PC-F and WS-F, two N₂O flux peaks were observed for mid–late July and mid–late October. Additionally, the flux rate of N₂O was lower in PC-F and WS-F than in F for most sampling times. The maximal flux rate of N₂O was 21.08 μg m⁻² h⁻¹ for WS-F on 22 July and 28.01 μg m⁻² h⁻¹ for PC-F on 8th October. In general, the mean flux rate of N₂O significantly decreased by 60.20% and 67.71% when using PC-F and WS-F in comparison to F, respectively (Table A1). Compared with F, the vegetable–fish co-culture significantly mitigated N₂O emissions during the study period.

3.2. Concentrations of CH₄ and N₂O in the Water and Bottom Soil

The concentration of CH₄ in the water and bottom soil did not significantly differ between the fish monoculture and the vegetable–fish co-culture treatments (Figure 3). The content of N₂O in the water significantly decreased by 24.39% and 44.41% when using PC-F and WS-F in comparison to F, respectively. This indicates that the vegetables primarily reduced the N₂O in the water.

3.3. Carbon and Nitrogen in the Water and Sediment

There were significant effects of the vegetable–fish co-culture on the content of NO₃⁻ in the water but no significant effects on the other parameters in the water and sediment (Figure 4 and Figure 5). The mean content of NO₃⁻ in the water was reduced by 23.67% and 30.08% when using PC-F and WS-F, respectively, as compared with F (Figure 4d).

3.4. Correlation Analysis

The Pearson’s correlation analysis showed that there was a significant positive correlation between the flux rate of CH₄ and the contents of NH₄⁺ and NO₃⁻ in the bottom soil of the fish monoculture treatment (

Table 1. Pearson’s correlation coefficient between the flux rates of CH₄ and N₂O and the parameters of the water and bottom soil.

		CH4			N2O
		F	PC-F	WS-F	F	PC-F	WS-F
Water	TOC	0.14	−0.26	−0.33	0.55 **	0.28	0.37
	TN	−0.01	−0.36	−0.36	0.10	−0.08	0.21
	NH4+	−0.27	−0.28	−0.22	0.12	−0.18	−0.14
	NO3−	−0.02	−0.23	−0.09	0.65 **	0.68 **	0.40
	NO2−	−0.23	−0.27	−0.26	0.86 **	0.24	0.56 *
	BOD	0.40	−0.30	0.09	−0.01	0.17	0.31
	COD	0.44	−0.13	−0.31	−0.002	0.40	−0.15
	pH	−0.20	0.58 *	0.43	−0.33	−0.53 *	−0.40
Soil	SOC	0.11	−0.44	0.17	−0.20	0.51	0.09
	DOC	0.27	0.20	−0.11	−0.22	0.001	−0.54
	MBC	−0.14	−0.22	−0.56	0.29	−0.15	0.59 *
	TN	−0.02	0.33	0.63 *	−0.31	−0.30	−0.35
	NH4+	0.71 **	0.01	0.08	0.10	0.07	−0.10
	NO3−	0.86 **	−0.18	0.12	0.43	0.50	−0.17
	DON	−0.26	0.09	0.65 *	0.08	−0.34	−0.19
	MBN	−0.09	−0.19	0.18	0.33	−0.09	0.24

* and ** indicate significant correlations at the 0.05 and 0.01 levels, respectively.

), while the flux rates of CH₄ in PC-F were positively correlated with the pH value in the water. As for WS-F, the flux rate of CH₄ was positively related to the content of TN and DON in the bottom soil. More water parameters showed a significant correlation with the flux rate of N₂O than CH₄. The contents of TOC, NO₃⁻, and NO₂⁻ in the water showed a significant correlation with the flux rate of the N₂O of F. As for PC-F, the flux rate of N₂O showed a significantly positive correlation with NO₃⁻ but a significantly negative correlation with the pH value. The flux rate of N₂O of WS-F showed a significantly positive correlation with the content of NO₂⁻ in the water and MBC in the bottom soil.

3.5. Nitrogen Utilization Efficiency, GWP, Yield, and GHGI

Table 2. N input, harvest, and utilization efficiency of different vegetable–fish co-culture systems.

Treatments	N Input (g)			N Harvest (g)		N Utilization (%)
	Feed	Fry	Vegetables	Fish	Vegetables
PC-F	150.83	7.44	1.10 a	52.61 a	12.47 a	43.89 a
WS-F	150.83	7.44	1.21 a	49.52 a	3.78 b	35.33 b
F	150.83	7.44		49.98 a		33.14 b

Different letters indicate statistical significance at the 0.05 level.

shows that the N harvest of the vegetables was significantly higher for PC-F than WS-F, while the N harvest of yellow catfish showed no significant difference among the three treatments. The utilization efficiency of PC-F was 43.89%, which was significantly higher than that of WS-F and F. These results indicate that the PC-F treatment was more effective in terms of nitrogen utilization.

The total CH₄ and N₂O emissions are shown in

Table 3. Total emissions, yields, and GHGI of the fish monoculture and vegetable–fish co-culture treatments.

Treatments	Total Emission			Yield (kg ha−1)		GHGI (kg CO2-eq kg−1)
	CH4 (kg ha−1)	N2O (kg ha−1)	GWP (kg CO2-eq ha−1)	Vegetables	Fish
PC-F	32.26 ± 4.26 a	0.31 ± 0.02 b	899.32 ± 99.73 b	67,577.38 ± 13,731.03 a	4372.17 ± 389.44 a	0.01 ± 0.01 c
WS-F	42.29 ± 2.55 a	0.25 ± 0.01 c	1132.61 ± 62.93 a	18,366.66 ± 5379.30 b	4230.49 ± 361.03 a	0.05 ± 0.01 b
F	35.12 ± 1.14 a	0.78 ± 0.03 a	1111.35 ± 34.35 a	-	4346.03 ± 76.64 a	0.26 ± 0.01 a

The yield of the vegetables and the fish was the fresh weight. Different letters indicate statistical significance at the 0.05 level.

. The results show that co-culturing with vegetables did not significantly influence the total CH₄ emissions during the study period, while it did significantly decrease the total N₂O emissions. The total N₂O emissions for PC-F and WS-F were reduced by 60.20% and 67.71%, respectively, compared with F. Additionally, the overall GWP of CH₄ and N₂O was significantly reduced by 19.08% when using PC-F compared to using F (Table 3). However, the overall GWP did not differ between WS-F and F due to the relatively higher CH₄ emissions from WS-F than F. The GHGIs were significantly reduced by 96.15% and 80.77% by PC-F and WS-F compared to F, respectively, although there was no significant difference in terms of fish yield among the three treatments. Owing to the higher yield in the PC-F treatment, the GHGI of PC-F was significantly lower than that of WS-F. These results show that co-culturing with PC was more effective in mitigating the GHGI than WS.

4. Discussion

4.1. Effects of the Vegetable–Crop Co-Culture on CH₄ Emissions

Previous studies have reported that floating aquatic plants could affect the CH₄ emissions from natural water bodies [23,24,25]. Either reduced or increased CH₄ emissions were observed for the floating aquatic-plants-growing area than the no-plant-growing area, which helped to develop an understanding of the impacts of these plants on O₂ recycling and carbon transformation processes in different aquatic systems [23,24]. However, the results of this study showed that co-culture with pak choi and water spinach using a floating bed did not affect the CH₄ emissions from the tanks (Table 3). This can be attributed to the difference in C and O₂ supply between the aquaculture pond and natural water bodies. Floating aquatic plants or vegetables may affect the CH₄ emissions from water bodies through two possible mechanisms. Firstly, the root exudates or sloughed issues of vegetables may provide substrate carbon for CH₄ production [26]. However, as in the aquaculture ponds, the sufficient C supply originating from residual feed and fish excretion may weaken the effects of vegetables in terms of CH₄ emissions [27]. Secondly, floating plants could hinder the exchange of O₂ from the atmosphere to water bodies [28], thus influencing CH₄ oxidation processes. As in aquaculture ponds, mechanical aeration may greatly weaken the effect of co-culture vegetables on O₂ recycling processes.

4.2. Effects of the Vegetable–Crop Co-Culture on N₂O Emissions

N₂O emissions in aquaculture are mainly derived from the processes of nitrification and denitrification, which are influenced by environmental factors, such as hydrophytes, the content of available N, pH, DO, and water temperature [4,29]. The results of this study show that two obvious peaks of N₂O were observed on 8th July and 14th October in F (Figure 2b), which are possibly attributable to the relatively high contents of NO₃⁻ and NO₂⁻ and the low contents of DO in the water (Figure A1). At the earlier stage, owing to the lack of vegetable cultivation, the high amount of inorganic N in the water and sediment was associated with a higher release of N₂O from the process of microbial nitrification and denitrification in the fish monoculture system [4]. At the late stage, the peak of N₂O emissions was mainly due to two reasons. On the one hand, the higher content of NO₃⁻ determined a sufficient substrate supply for producing N₂O. On the other hand, the concentration of DO on 14th October was lower because of continuous rainy days, the deterioration of the water, and an increase in oxygen consumption for organic matter decomposition. Therefore, high levels of available substrate nitrogen with a low oxygen supply may prompt N₂O production [11,16].

Co-culture with pak choi and water spinach significantly reduced N₂O emissions (Figure 2 and Table 3), which is consistent with the results of previous co-culture systems with different crop plants [15,30]. Li et al. [15] reported that the total amount of N₂O emissions was reduced by 85.60% and 108.30% for a rice–catfish co-culture pond and a rice–shrimp co-culture pond, respectively, compared with that of a monoculture pond. Similar results were presented by Hu et al. [30], who found that a rice–fish co-culture significantly mitigated the total N₂O emissions by 56.20% in a yellow catfish pond. These studies indicate that co-culture with vegetables or other crop plants was beneficial to the reuse of residual nitrogen in water and helped to reduce the substrate nitrogen for N₂O production. This was supported by the significant reduction in NO₃⁻ in the water containing PC-F and WS-F (Figure 4d). Previous studies have also shown that leafy vegetables tend to assimilate and store much more nitrate [31,32], which provides evidence for the mitigation of N₂O emissions. However, it is still difficult to determine whether the mitigation efficiency of pak choi and water spinach for N₂O emissions is better than the other crops tested in previous studies [11,15,20]. This is because the structure of the co-culture system, fish species, and the flux rate of N₂O were different in these studies. As in this study, water spinach led to a greater reduction in N₂O emissions than pak choi (Table 3). However, the total amount of nitrogen absorbed by water spinach was less than that of pak choi. This was possibly because the rhizosphere effect of water spinach on nitrogen transformation may be beneficial in the reduction of N₂O emissions [33,34]. Water spinach is a stem vegetable and has stronger aerenchyma than pak choi. The root secretion of O₂ by water spinach may prompt the reduction of NO₃⁻ and N₂O eventually to dinitrogen, thus reducing N₂O emissions [4,33]. The mitigation of N₂O emissions by vegetable cultivation may be attributed to the contents of the substrates. Hence, to better reveal the mechanism of vegetable cultivation in mitigating N₂O emissions, our research will further focus on the changes in microbial functional genes involved in the processes of nitrification and denitrification in vegetable–fish co-culture systems.

4.3. The Balance of Yield and GHG Emissions

The area-scaled GWP values of PC-F and WP-F in this study were 899.32 and 1132.61 kg CO₂-eq ha⁻¹, respectively, which are higher than the rice–fish co-culture with a stocking density of 150,000 fingerlings ha⁻¹ (RF15), lower than that with a stocking density of 450,000 fingerlings ha⁻¹ (RF45), and similar to the RF30 in aquaculture ponds [30]. Compared with the monoculture treatment, the GWP of PC-F was significantly reduced by 19.08%, which may be attributed to the lower N₂O emissions (Table 3). The effect of the vegetable–fish co-culture on GHGI is determined by the overall impact of vegetable–fish co-cultures on the total emissions of CH₄ and N₂O and the yield of vegetables and fish. The GHGI results of PC-F and WP-F in this study were 0.01 and 0.05 kg CO₂-eq kg⁻¹, respectively, which are lower than the results of a previous study on rice–fish co-cultures [30]. The results indicated that co-culture with vegetables could mitigate GHGI emissions. This is probably because of the higher yield of vegetables. Furthermore, no other fertilizer was used in the co-culture system; thus, the vegetable–fish co-culture system might be an ideal model to mitigate GHG emissions and improve the yield from aquaculture ponds. Additionally, the GHGI of PC-F was significantly lower than that of WS-F, indicating that the PC-F co-culture system was more effective in the mitigation of GHG and the enhancement of the yield. This study only determined two vegetables’ effects on GHG emissions. Leafy vegetables, such as lettuce, spinach, cabbage, and kale with a floating bed appear to be suitable plants for co-culture with fish in aquaculture ponds [34]. Further studies should be undertaken to screen the effective species of vegetables on GHG emissions and make clear the underlying mechanisms of GHG mitigations.

5. Conclusions

Co-culture with pak choi and water spinach significantly reduced N₂O emissions, but it did not significantly affect the CH₄ emissions from yellow catfish aquaculture. Pak choi and water spinach cultivation only significantly reduced the content of N₂O and NO₃⁻ in the water. Furthermore, vegetable cultivation significantly improved food yield and N utilization. Moreover, co-culture with pak choi and water spinach reduced GHGI due to the additional harvesting of vegetables compared with yellow catfish monoculture. The results indicated that both these vegetables are ideal for inclusion in a co-culture with fish to reduce N₂O emissions with greater food yields.